Pillars of Change in Music Therapy and the Extended Evolutionary Meta Model (EEMM)

Music Therapists do some amazing work across the country, and many of them can recount clients with whom they’ve experienced significant, profound change in their presentations. Having experienced this sense of change from early on in training, it was initially hard for me to capture. There are a handful of music therapy assessments out there that aim to track this change. But what I found as I grew more experienced was that a lot of these assessments were significantly diagnostic in nature, and I feel this is evidenced by the development of lots of population-specific assessments, or at least the inconsistent application of assessment tools. Indeed, having interviewed many creative arts therapists over the past year, one of my favourite questions has been about outcome tools and assessment tools. Rather than wanting to hear they have used anything in particular, I want to hear that they have grappled with the process of using one; that they have critically evaluated the pros and cons of them.

Almost without exception, the feedback is yes, I have used them, but they don’t “feel right”. As a result some continue to use them despite this feeling, others drop them entirely.

I’m not surprised by this when I hear it, but it’s the exact process I want to know that clinicians I hire have been through. What is telling, for me, is that no music therapy-specific, trans-diagnostic assessment has found its way into standard practice other than qualitative and goals-based reporting - which I think should always be in place in music therapy, anyway. Having thought on this some more, I think that this comes down to the idea of what it is that we are measuring, and what the implications are of measuring in that way.

Topography change

The main form that I think comes most naturally to measurement is the topography of our sessions. I would define topography as Process-Based Therapy would (more on that later): the immediate, tangible, form and shape of a person’s behaviour which leads people — therapists included — to draw conclusions around what is happening for that person.

This makes sense as a form of change that we focus on, primarily because it is present to us and indisputable. These forms of change can be summarised as:

1. Duration: how long do they do it for?
2. Repetition: is the expression/behaviour repeated, and if so, how many times?
3. Latency: how delayed is the expression/behaviour after the antecedent?
4. Amplitude: how much of the expression/behaviour takes place, in the domains of speed, volume/strength, or quantity of materials used?

Tracking the topography should take place in all assessment. It forms the basis of the facts that are taking place before you. The risk is that, because it’s easy to monitor, that it becomes all that we monitor. And whilst we may want to increase the duration of a musical interaction in order to make a correlation between a client and their sense-of-Self, it is arguably a false equivalence to assume that increased duration is, therefore, increased sense-of-anything.

Essentially, topographical change risks being category errors if measured in isolation, but is the easiest measure to take. Here’s why.

The topography trap

Within Contextual Behavioural Science, they emphasise the importance of functional equivalence and difference in order to show why behaviour needs to be understood within the context in which it takes place, both broad and specific. They point to this as why simple behavioural analysis, as well as psychotherapy too, can be substantially limited.

Consider the following scenarios:

• A: A client plays a steady pulse for five minutes.
• B: A client plays a violent cymbal crash.
• C: A client sits in total silence.

Now consider how those scenarios might map onto the following functions:

• X: Uncertainty or anxiety about what to do next.
• Y: Somatic grounding and self-soothing.
• Z: Trying to keep interpersonal distance.

We can sit here and make topographical assumptions based on ideas we might traditionally expect to see in our work:

Scenario A might well be led to us assuming Function X. They may not seem to be responding to us, leaving us to label the client as “stuck” or “perseverative.”

Equally, Scenario B can be viewed as topographically opposite to Scenario C. We may describe Scenario B as being “fragmented” and C as being “avoidant”, or Function X.

But it is probably obvious by now that it isn’t that simple. Scenario A and C could just as easily be Function Y: seeking to self-soothe in order to maintain regulation. The client may care deeply about remaining regulated in order to preserve the relationship. Therefore, it isn’t so simple as to say that sitting in total silence is avoidance, or anxiety. They may have functional equivalence. So we can see that opposite actions may well serve the same function.

The exciting bit is when you open the possibility that the same scenario can serve for opposite functions. Let’s take Scenario A: a client playing a steady pulse for five minutes. A client with a low sense of self-efficacy might well be uncertain or anxious about what to do next (Function X), which might lead to a tendency to be pliant or prescriptive in their interaction. However, it could well be that they are deeply invested in what they are playing, and finding it to be Function Y; a highly purposeful and intended form of music-making. The same is said of Scenario A and C; topographically opposite - lots of playing, vs. no playing. And yet, both could serve that role of self-soothing, and by assuming that sitting in silence for a whole session is indicative of not connecting, we could be missing the real communication.

So what can we measure?

This presents a problem. If we can’t necessarily trust topographical measurements to tell us anything about what’s going on, then what can we rely on? We are back in the realm of purely qualitative to see us through.

Well, not quite. When you consider a map, you need to know the landscape in order to know where to go. However, you also need a way to orient yourself within that map, something which gives you a sense of where you are in amongst the coordinates. This is where we can lean on new forms of change, which are borrowed from process-based therapy.

Process-based therapy (pioneered by Steven C. Hayes and Stefan G. Hofmann) does away with the traditional medical-model assumption that symptoms cluster into diagnostic boxes requiring prescriptive protocols. Instead, PBT asks a fundamentally idiographic question: what core biopsychosocial processes need to be targeted with this specific person, in this specific context, to help them move towards their values? Rather than viewing a client through the lens of a static disorder, PBT conceptualises psychological life as a dynamic, interconnected network of modifiable processes that evolve over time.

Crucially, process-based therapy hinges on the understanding that diagnostic-based models are not enough to capture the complexity of lived experience, leaning into the contextual side of its history in order to understand that people and clients, when we see them, are a coordinate in space and time in a grand network map of their life so far. They, therefore, rely on the universal understanding of how life develops and moves: evolution.

Within process-based therapy, they outline 4 patterns of change, across 8 systems split into 6 dimensions and 2 levels, identifying both adaptive and maladaptive processes which are taking place. This is called the Extended Evolutionary Meta Model, based on the Extended Evolutionary Synthesis in biology. We will look at the patterns of change below, but the systems are:

The dimensions:

• Affective: how emotions and physiological feelings are modulated.
• Cognitive: beliefs, appraisals, cognitive framing, and symbolic reasoning.
• Attentional: how consciousness is directed.
• Self: the perspective from which one experiences.
• Motivational: what drives behaviour?
• Overt Behaviour: observable physical action, motor habits, and communication.

The levels:

• Biophysiological,
• Sociocultural.

An adaptive process is one which is flexible and context-appropriate. It is identified as being one which moves people towards what matters to them, often linked to “workable behaviour”. Maladaptive processes are inflexible and context-insensitive, likely formed at a time when the individual was under stress and it proved successful in the immediate relief and change in the environment. For instance, a young person with PTSD who attacks a member of staff when they get too close may have had to defend themselves in the past, which was “workable” within that specific context. It grows to become maladaptive because it no longer appears within the “workable” context, instead appearing generalised and trans-contextual.

A dimension is an aspect of our psychology through which we develop our sense of Self. These come from within ourselves, and are grounded in the primary targets of established evidence-based therapies (CBT, ACT, DBT, and Psychodynamic therapy). The boundary levels are aspects either of our nature or environment in which our psychological processes are nested. It’s important to have the boundary levels, so that we understand that we are inherently linked with both our environment and our nature.

Process change

At the heart of the EEMM is Donald Campbell’s concept of Universal Darwinism: the evolution through the same core algorithm which applies to all adaptive systems (essentially, life):

Variation × Selection × Retention within Context

PBT adapted this to explain psychopathology and psychological growth:

1. Variation: a healthy psychological system generates a wide, flexible repertoire of responses (behavioural options, thoughts, reactions), with difficulties stemming from severe rigidity (avoidance, compulsivity, rumination) or chaotic, non-functional variation (severe dysregulated impulsivity).
2. Selection: contextual pressures select certain thoughts, emotions, or behaviours over others. Difficulties come about when selection is driven by immediate short-term relief (experiential avoidance) rather than long-term value or consideration.
3. Retention: selected patterns are stabilised, encoded, and repeated until they form persistent habits, cognitive schemas, or biological feedback loops.
4. Context: these things never happen in a vacuum. A behaviour is adaptive when it is relative to the environment, social, and physiological context in which it occurs.

These four pillars of change track how an individual’s behavioural and relational repertoire adapts, rather than the musical product they leave behind. This is where we get back to the map - the evolutionary pillars of change make up the coordinates which we overlay on a map consisting of topography, therefore making sense of ourselves in relation to the topography.

It is important that we learn to notice both because a map without ourselves in relation to it is just an image, and coordinates without a map means we are guessing where we are going. Adapted for behaviour in a music therapy room, you might begin to notice these changes:

1. Variation: how much does the client change what they are doing, with what’s available to them?
2. Selection: what are the patterns for how the client makes choices (e.g., self-led, contextually appropriate, prescriptive)?
3. Retention: does the client retain and consolidate an adaptive emotional or relational pattern once it has been selected in the music?
4. Context: does the client fit their behaviour to the musical context of which they are part?

We can measure process-based change through topographical change, too, but importantly the ability to measure musical interaction and interpersonal interaction through both topography and process means that we can create a form of measurement which is not founded purely on ‘more music’ or ‘less music,’ ‘more interaction’ or ‘different interaction’ but is instead based on aiming for interpersonal, musical interaction which responds flexibly to the context. In order to do that, we have to keep measuring the topography, but we must treat it as raw observational data rather than any clinical evidence of wellness. Topography gives us the what — the measurable volume, tempo, duration, and silence occurring in the physical space. Process gives us the how and the why — whether that behaviour is being deployed rigidly or flexibly in response to the relational landscape. Without topography, our assessment lacks an anchor in the music; without process, our assessment lacks an anchor in the human being.

Trans-diagnostic measurement and contextual generalising

Measuring in this way means we can truly access a trans-diagnostic form of assessing in music therapy. Topographical measurements assume there is “an amount” of playing, or “a way” of playing which we are aiming for. And sometimes, this might be true; it is a totally valid goal to want to support someone to play more. However, a formalised assessment based on this will run into challenges immediately depending on the cohort. Somebody with a learning disability, for example, may simply never have short latency between antecedent and response. Or, somebody with tics or dystonia may always play in a way which is hard to distinguish from impulsivity challenges. Topographical tools naturally fracture along diagnostic lines. The moment a client’s physical presentation disrupts standard expectations of topographical measurement, the tool fails unless it is calibrated to a specific diagnosis.

However, the four pillars of evolutionary change are intentionally universal. They apply to people regardless of their age, diagnosis, or presentation, and are all about moving people towards living as meaningfully or flexibly as they can. There is no ‘correct’ amount of flexibility. When we account for process-based change, a client who experiences difficulties with motor control can still be monitored for the shifting in their variation, or selection, whilst we account for the topographical tendencies that come from their diagnosis. Equally, a young person with a learning disability can be assessed against how adaptive their retention of a musical interaction is, rather than how quickly they respond to a musical antecedent.

And for all clients, this offers an opportunity to build on contextual generalising. Process-based skills — such as flexible selection and situational sensitivity — are functionally transferable. By contrast, purely topographical change is often context-limited; a client might learn to play rhythmically or regulate through music in the presence of the therapist, but because that behaviour risks being trained as a surface form rather than an adaptive process, it collapses when they return to the classroom. I would propose that, were we taught to notice and train the process-based mechanics, we may notice that therapeutic change generalises more readily into new contexts and environments.

Moving us forwards

If we want assessment in music therapy that captures the nuance of our work without abandoning scientific rigour, and without having to overly rely on self-reporting scales (which are often diagnostic in nature themselves, and inaccessible to many of the client groups we work with), we have to start tracking living, functional processes.

I’ve been exploring this bridge between music therapy, process-based therapy, and relational frame theory — and how symbolic, relational learning operates in the music room — in a forthcoming paper. I’ll be speaking about this exact intersection at the BAMT conference this November. If you’re grappling with these questions or are curious about how RFT underpins process-based music therapy, do attend my session on Sunday 22nd November at 4:30pm. Alternatively, I’m always happy to speak about these ideas if anybody wants to reach out for a discussion.

hello@thatmusictherapylot.uk